jason-easyazz/agent-zero-research
skills/agent-zero-research/SKILL.md
Delegate deep research, multi-source analysis, and comparison tasks to Agent Zero. Use when the user asks to research a topic, compare options, investigate something, or wants a 'deep dive' with cited sources.
$ install --global
skillsauthnpx skillsauth add jason-easyazz/zoe-ai-assistant agent-zero-researchInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: May 26, 2026, 7:47 AM82.3s1 file scanned
SKILL.md
- name:
- agent-zero-research
- description:
- Delegate deep research, multi-source analysis, and comparison tasks to Agent Zero. Use when the user asks to research a topic, compare options, investigate something, or wants a 'deep dive' with cited sources.
- version:
- 1.0.0
- author:
- zoe-team
- api_only:
- true
- priority:
- 5
Agent Zero Research
When to Use
The user wants deep research, multi-source analysis, comparison studies, or any task requiring web browsing, document reading, or synthesizing information. Triggers include "research", "compare", "deep dive", "pros and cons", or "look into".
Workflow
- Identify the task type — research (open-ended query), comparison (two+ options), or analysis (structured evaluation)
- Inform the user — tell them research is underway and may take 30–60 seconds
- Submit to Agent Zero — use the appropriate endpoint (see below)
- Poll for completion — use the submitted task ID with
GET /api/agent-zero/status/<task_id>until the task finishes - Present findings — format results clearly: use tables for comparisons, bullet points for research summaries, and always cite sources when available
API Endpoints
Submit a research task
POST /api/agent-zero/research
{
"query": "Best solar panels for residential use in 2026",
"depth": "thorough",
"max_sources": 10
}
Submit a general task (comparisons, analysis)
POST /api/agent-zero/task
{
"task": "Compare Tesla Powerwall vs BYD battery for home storage",
"context": "User is considering home battery storage"
}
Check task status
GET /api/agent-zero/status/<task_id>
Returns the current state (pending, running, completed, failed) and results when complete.
Example
User: "Compare Tesla Powerwall vs BYD battery for my home"
Steps:
- Identify: comparison task with two items
- Respond: "Let me research that for you — this may take about a minute."
- Submit:
POST /api/agent-zero/taskwith task and context - Poll:
GET /api/agent-zero/status/<task_id>untilcompleted - Present: table with capacity, price, warranty, pros/cons for each option, with source links
Error Handling
- Task timeout (>120s): Inform the user the research is taking longer than expected and offer to retry or narrow the scope
- Task failed: Check the error from the status endpoint and relay a user-friendly message (e.g., "Agent Zero couldn't complete the research — want to try a more specific query?")
- No results: Suggest the user rephrase or narrow the research topic
Formatting Guidelines
- Research results: Bullet-point summaries with source citations
- Comparisons: Side-by-side table with key attributes
- Analysis: Numbered findings with supporting evidence
Related Skills
jason-easyazz/autoresearch-engineer
tools
VerifiedTrustedCommunity
Run Karpathy-style fixed-budget optimization loops for one approved asset and one objective score. Use when the user asks to optimize, auto-research, run overnight experiments, improve a prompt/tool/copy/config, or turn 'is it good?' into a number.
3SKILL.mdUpdated Jun 11, 2026
jason-easyazz/skills/web-search
development
VerifiedTrustedCommunity
# Web Search Skill ## When to Use Use this skill when the user wants information that requires looking at the live web. ## Trigger Conditions **Use `web_search` (fast, ~3-5s) when:** - Single-source fact lookup: news, exchange rates, sports scores, weather, stock prices - One specific product at one named retailer: "what does Bunnings charge for X" - Simple factual question answerable from one good search result **Use `deep_web_research` (~60s) when:** - ANY mention of location / "near me"
3SKILL.mdUpdated May 25, 2026
jason-easyazz/skills/touch-panel
tools
VerifiedTrustedCommunity
# Touch Panel Skill Zoe drives physical kiosk panels via `panel_*` MCP tools and can SSH into them for diagnostics and repair. ## When to use this skill - User asks about the touch screen / panel / kiosk - Diagnosing why the panel is blank, frozen, or showing wrong content - Restarting, updating, or re-provisioning a panel - Controlling what shows on the panel (navigate, announce, smart-home overlay, etc.) - Registering a new panel or managing panel tokens --- ## Current hardware (productio
3SKILL.mdUpdated May 25, 2026
jason-easyazz/skills/openclaw/zoe-widget-builder
tools
VerifiedTrustedCommunity
# zoe-widget-builder Build new dashboard widgets for Zoe's touch panel and desktop UI. ## Trigger conditions This skill activates when the system message begins with `[ZOE_SELF_BUILD: widget]`. ## Prerequisites - Caller must have admin role. Check via `zoe_self_capabilities` tool (role field). If not admin, reply: "Widget building requires admin access." - Do NOT build if the widget already exists (`zoe_self_capabilities` returns it in existing_widgets). ## Step-by-step workflow ### 1. Chec
3SKILL.mdUpdated May 16, 2026